Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

Children of Grainger: Leather Instruments for Free Music

Stuart Favilla Joanne Cannon Bent Leather Band Bent Leather Band Hybrid Initiative Hybrid Initiative ICEPA ICEPA Victoria University, Melbourne Victoria University, Melbourne 61+3 97301026 61+3 97301026 [email protected] [email protected] ABSTRACT We first discovered Grainger’s Free Music back in the mid Grainger’s Free Music remains a rich source of discovery for 1980s, in the library at LaTrobe University. From there we contemporary Australian musicians. Free Music represents a found the Grainger museum; which has had a long significant departure point for electronic musicians and association with experimental . Finally we instrument makers searching for new musical language, form began our collaboration with Garry Greenwood and through and expression. This paper presents research undertaken by the exhibition of this work at the Grainger Museum, created the Bent Leather Band exploring Grainger’s Free Music ideas the opportunity to work there with our Blisters project. within a 21st century music-making context embracing live improvisation, instrument and software design. Research Grainger’s Free Music comprises writings, recordings and the outcomes presented in this paper includes a range of creative actual machines he created. “Music beyond the traditional works; meta-serpent wind controllers, the 4th generation of constraints of pitch and rhythm” [4], was developed through the construction of many bizarre instrumental experiments the light-harp controller, new MAX-based software engines th for processing, control-modes and strategies for the and prototypes including; 6 tone tuned pianos fitted with instruments and music including Bent Leather Band’s latest player systems, air-pump powered reed organs capable of collection of works “Children of Grainger”. This paper fine controlled portamenti, and large machines; such as the discusses technical issues confronting the contemporary kangaroo pouch; which allowed the pitch of up to four electronic instrument builder and presents Bent Leather electronic valve oscillators to be played by score of cut Band’s aim to develop playable instruments. cardboard and paper rolls. Grainger’s explains his Free Music as “music using gliding 1. FREE MUSIC tones and irregular rhythms” throughout the 1951 recordings Since late 2003, Grainger’s Free Music has been revisited by of his experiments and instruments. The 1938 manifesto [4] a number of Australian composers and experimental explains his desire to liberate or free sound from the musicians. In all cases the Grainger Museum audio collection constraints of conventional pitch and rhythm. has revealed a surprising amount of interesting material that has redefined our previous notions of Grainger’s Free Music “Existing conventional music (whether "classical"; or experimental depth and rigor. Warren Burt’s work covers the popular) is tied down by set scales, a tyrannical (whether history of the Free Music experiments [1] and presents an metrical or irregular) rhythmic pulse that holds the whole appendix of the audio collection. His work rebuilding tonal fabric in a vice-like grasp and a set of harmonic Grainger’s unfinished Electric Eye Tone Tool; a seven part procedures (whether key-bound or atonal) that are merely Free Music player machine, was supported by the ABC habits, and certainly do not deserve to be called laws.” listening room and presented collaborative works for the [Grainger, 1938] Electric Eye by Tristram Carey, Catherine Schieve, Wang Zheng Ting and Warren himself. The Blisters ensemble; an This manifesto also reveals a desire to bypass the role of a ensemble of Australian improviser/instrument builders performer or interpreter of his music. It directly follows including; Jon Rose, Rainer Linz, Tom Fryer, Joanne Cannon Grainger’s experiences composing for and rehearsing and myself, were also commissioned by the ABC listening ensembles. His attempts to have his Free Music room to investigate Grainger’s Free Music legacy and this compositions played by musicians on never created a radiophonic work “Skeleton in the Museum”, which achieved results to his satisfaction [5]. It is doubtful that was selected for the 2004 International Karl Szcuka Pries. Grainger knew about categorical perception [17] and perceptual limits on the human ear and how it would affect The Bent Leather Band has continued to work on Grainger. the performance of glides using theremins. Grainger was We were surprised to discover the diversity of Grainger’s already using technologies and the potential of experiments. The breadth of the audio collection was a stark piano roll devices became the hub of his Free Music contrast to the musical education we received in Australia; activities. which sorely neglected Grainger let alone his experiments. Permission to make digital or hard copies of all or part of this work for As improvising live ensemble musicians, we had to consider personal or classroom use is granted without fee provided that copies how much Grainger’s preference for paper-roll [player] are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy sequencing was going to influence the formation of our otherwise, or republish, to post on servers or to redistribute to lists, music. As improvisers, we are not interested in a paper-roll or requires prior specific permission and/or a fee. sequencing technique. But perhaps we share with Grainger a Nime’06, June 4-8, 2006, Paris, France common desire to eliminate the “proverbial middleman” or Copyright remains with the author(s). ©copyright bent leather band 2006 interpreter (Cross, 1976, in Drefuss interview [6]). We do

370 Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

know how impressed Grainger was with improvised music Ulyate, [18]. His ten commandments of interactivity require generally. He brought the Ellington Band into one of his “no expertise”, “no thinking” and measure quality of control composition classes in New York [5] and Grainger’s own top from immediate reactions of players or participants. ten system rated Rarotongan improvised polyphony 3rd; well above Debussy and even Bach [7]. Therefore, as improvisers, Other approaches proposed by Mulder [9], develop new we embrace Free Music as an opportunity to escape the rigid instruments to existing human motor skills, rather than harmonic constraints of traditional pitch systems and also as a requiring the musician years of commitment to developing departure point for the development of new specialised new skills. Pedro Rebelo’s work applies the media theory of musical instruments. prosthesis to instruments [15]. Physical modeling is used in an intervention of an acoustic sound to mimic, extend or fulfill a potential of the body [acoustic instrument]. For Rebelo the player’s intention and the instrument [which he defines as a point of resistance] constitutes an acoustic threshold.

Other interesting areas of research include: The continuing development of controllers modeled from existing acoustic instruments such as; Cormac Cannon’s EpipE Uilleann Pipes [2] and Diana Young’s HyperPuja [18]. Completely new novel instruments designed for a specific form of synthesis or sound generation; including Sile O’Modráin and Georg Essl’s PebbleBox and CrumbleBag controllers for [12]; and Blockjam [11], a polyrhythmic sequencer interface that forms a series of interconnecting block switches. The switch’s function is displayed by an LED panel and can change throughout an interaction or piece.

Figure 1. Detail of Grainger’s Kangaroo Pouch Machine Something that seems to be lacking from the field overall is a [Courtesy Grainger Musuem] development of new original music through new instruments/interfaces and although there are many new 2. NEW INSTRUMENTS contributions made to the field in the form of new instrument The field of new interfaces for musical expression continues prototypes, very few of these prototypes are developed to the to expand. Musicians are offered increasing access to new next generation. The field has also responded to the rise of technologies that can develop new instruments. Network sound design over music. protocols such as OSC, are beginning to purge the old MIDI language through a range of new interfaces from Kroonde Our work has been primarily concerned and Gluion. Micro-electronics internet groups such as the with skilled ensemble performance of new sounds. Our broad MIDI Box network run by Törsten Klose, have made research aims have been to create new music, performance available cheap MIDI circuits and PIC chip-software; and ensemble techniques and new instruments. So far our allowing musicians the chance to construct their own work has developed in order to embrace specific musical customizable interfaces. Novel controllers, mixers and DJ languages. Over the past five years, our language has spatial sensor interfaces for music are available straight off specialized in beat-less, microtonal and gliding forms of the shelf in music stores. sonic expression. In essence we have been playing a digital form of Grainger’s Free Music. A constant proliferation of theoretical literature regarding the development of musical instruments has also flourished over 3. EVOLUTION OF PLAYABILITY the past five years. Amongst this proliferation are ideas that Our idea of a playable instrument is one that essentially does challenge the relevance of traditional notions of music not limit or inhibit the development of skill. The key is a performance such as whether the role of virtuoso balance between the instruments’ expressive potential, performance is valid? Choi[3], and Paine[13]: and the responsiveness, quality of feedback, embodiment of the blurring of the traditional roles of composer, performer and sound and the instruments’ ability to provide the player with listener. an intuitive understanding about the music being played. The instruments we were going to build had to suit the music we Twenty years ago Jeff Pressings imagined a super-instrument played and also work well together in ensemble. We defined [14]. It posited a human limit of up to ten independent this as playable, meaning degrees of freedom and provided the player with multiple channels of quality sensory feedback. Controllers were  Expressive expected to develop high resolutions, scanning rates and  Responsive sensitivities capable of performing very fine expressive  Versatile in solo and ensemble performance control of music. Pressing’s work defined ten fundamental  Visceral [naturalness, appropriateness, good visual issues relating to the design and construction of new feedback] interfaces.  Palpable [allowing for skill development, an instrument you can practice for hours] Although Pressing’s instrument has been achieved in part,  Inspiring [intuitive, revealing new things to the much of today’s work lies in the domain of instruments for player] public interaction. In stark contrast to Pressing is the work of  An instrument that has a definitive sound or character

371 Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

The focus on playability was intended to unify all aspects of response time of light-sensors [less than 2msec] and makes controller interface design across as many possible sensing beams playable of up to 200 MIDI notes a second. disciplines, e.g. cybernetics, HCI, ergonomics, gesture This means that unlike conventional keyboards and other research and skill-development. Specific areas we have controllers, the Light-harp is capable of performing focused our preliminary research on have included tactile, extremely dense and interesting textures not to mention haptile control and expert skill development Shakel [16], the glissandi. It is well suited to the performance of equally combination of dominant [attack based] gestures with tempered microtonal tuning systems such as the sixth tone ancillary or [modifying] gestures and the potential limitations tuning system used by Grainger to approximate glides and of bimanual control [8]. perform “loud unisons”, [tremelos] with his own butterfly The goals of the project were as follows: piano. Within the Free Music project the Light-harp has referenced the butterfly piano by using piano samples as a 1. To build an ensemble of new playable electronic source timbre for all sound creation. instruments 2. Develop a new improvised music and ensemble 3. Build prototypes and develop them into mature aesthetic instruments 4. Explore the language of ’s Free Music

The overall aim of the research was to create successful instruments, their playing techniques and ensemble music simultaneously. Our process was reflexive, sometimes beginning with a sound or process and then finding physical gestures that could effectively control them. Sometimes gestures discovered their own sounds. We investigated playability through the development of two distinct instrument projects; one investigating the potential of light sensors to trace virtual strings for a musician to play [the Light-harp]; and another project investigating the potential of live signal processing control of double reed [the Meta- serpents].

4. LIGHT-HARP The Light-harp uses spotlights and lasers to trace virtual strings through space. The instrument is a MIDI controller and was originally built in collaboration with David Brown [a Figure 3. Light-harp and shakuhachi maker] and Robin Whittle [a notable Dimensions 164 x 64 x 29cm instrument developer and designer]. The experience of building two previous instruments has brought about changes to the instruments dimension and shape. The neck now supports a scalloped tactile playing surface so the player can feel the sensors sitting under the fingertips. The curvature of the neck has been increased making the instrument’s dimension more compact and additionally, the ancillary controllers have been grouped in accordance with bimanual control [8], emphasizing a strong- handed [in this case right-handed] role for leading attack gestures against a left handed passive modification role. Whenever possible, a breath controller is used to control dynamics or attacks. Figure 2. Light-Harp Ancillary Controllers The aesthetic design of the instrument merges elements from After earlier models were built, using wood, fiberglass and Indian music and 1930’s valve radio equipment. Bakelite and steel as construction materials, the current Light-harp was French polished controller knobs are set against flat polished made from leather by the talented Tasmanian leather artist, leather panels. The Indian elements include the dragon [yali] Garry Greenwood. This version supports an extensive array headpiece, human physiology of the pelvis [instrument base], of controllers. These include an active electro-magnetic spine & vertebrae [neck and sensors], lotus flower [the whammy-bar, a two-dimensional bamboo whammy-bar, two tailpiece], and the fluted trumpet end. With the exception of large wheels, breath control and two touch-sensitive strips. It metal control panels, a strip of supporting metal and wooden is usually played with up to 5 independent dimensions of pieces supporting the base, the instrument is constructed freedom. It is also boosted with a control panel of 16 entirely from leather. assignable pots for synthesis parameter control. The instrument controls , software-synthesisers and 5. SERPENTS signal processing. The evolution of the meta-instrument controllers began with the sensor modification of simple double reed instruments. The Light-harp's specialized hardware allows for the Joanne Cannon, a bassoonist and Australia’s chief threshold attenuation of light-sensors. This reduces the protagonist for the creation of an Australian electric bassoon,

372 Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

wanted to transport her reed playing into a signal-processing polished panels and an Indian theme of a lotus or orchid environment. The first prototype instrument used force design. sensitive resistors and a passive magnetic proximity sensor to track the spatial position of the instrument’s bell. This instrument was interfaced via a MIDI control circuit to a laptop running Max which in turn controlled a number of effects units. The musical language we developed for this instrument made heavy use of delays, which we used to create additional parts. These techniques required fine control of times and more controllers were desired to independently control the multiple audio streams. The major drawback of this instrument proved to be its limited tonal production. This led to the idea of making long tubes with open holes.

The second prototype instrument we built in collaboration Figure 5. Detail of force-sensitive resistors with instrument leather instrument maker Garry Greenwood. Contra-monster, 2005 The Serpentine-bassoon is a leather meta-bassoon, with a 2.4meter conical bore. The instrument has eight open holes; 6. FREE MUSIC MAPPINGS which can be used to play pitches or closed with stoppers Software is a necessary part of our process and a laptop is the allowing for sensors to be played instead. This instrument host computer for practical reasons. The Laptop is produced a variety of timbres reminiscent of bassoons and effectively the live effects studio and Max software allows all horns. Two contact condenser were used to pick of our sensors to be mapped to all the parameters we use to up a large variety of sounds and the signal was processed play our music. Our mappings are fixed, not dynamic, but we using MaxMSP via a Digi002. Dials were added for fine will usually switch between several mappings during a delay time and other parameter control and three force performance. The instruments however, constitute the sensitive resistors were used to control dynamic features of interface between the musician and software with the laptop the signal processing such as acoustic or delay feedback etc. remaining off. Sensor mappings consists of a number of process stages including sensor adjustment [rectification], rescaling, processing [including averaging/interpolation of data]; and finally the mapping and tweaking of a specific parameter of synthesis or signal processing.

Figure 4. Serpentine-bassoon Controllers attached Figure 6. Joystick, Contra-monster The third instrument, dubbed Contra-monster, has a 3.6meter conical bore and was built for maximum signal processing Signal processing and synthesis techniques have developed control. It has two built in condenser microphones, and 15 from experiments using delay with modulating or playable controllers including, three dials, one fine tuning dial, one delay-time. The technique is commonly associated with fader, two joysticks and six small force sensitive pads; in the echoes. However, if the feedback of the signal and the delay place of finger holes. The sensors have been positioned time can be accurately controlled, tones and independent ergonomically for ease and effectiveness of use and the lines can be achieved in a myriad of ways. interface was completed with a small built in display for the performer. The Contra-monster is capable of ten Other associated signal processing techniques explored so far simultaneous degrees of freedom. have included; pitch-shifting [coarse, tuned, continuous or modulated], extensive control of delays [to create pitched The current instrument was built around a MIDIBox Plus PIC feedback tones], distortion [overdriven or boosted , controller that was redesigned to make the circuit board ring-modulation, noise, clipping and unstitched wavelets], smaller. A small panel of push buttons allows for the and granular treatments [streams, clouds, pitch-shifting, instruments controller mode to be changed allowing for over distortion, prolongation, and accenting], this list is not 760 possible assignments for the MIDI controller signals. exhaustive. We developed processing techniques in MSP but also had great success hacking groups of existing VST format The aesthetics of the serpents combine the same elements plug-ins including those found in cycling’s Pluggo suite and adopted by the Light-harp. The Serpentine-bassoon was other freeware plugins. made as a direct relative to the Light-harp using the same color scheme and leather dyes. The Contra-monster’s visual Throughout the project we were conscious of providing each aesthetic combines 1930s Bakelite radio dials, French- instrument with its own character. This was achieved by

373 Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

limiting the source timbres. In the case of the serpents their does not necessarily mean realistic. Nor do we subscribe to source is the sounds created by the double reed anyway, but the research areas of audible gesture or universal musical the Light-Harp, a controller can play a huge gesture in relation to our work. We believe for example, that variety of sounds. In the theme of Grainger’s microtonal the tiniest movement of a fingertip is entitled to make the butterfly piano experiments, we limited the Light-Harp to hugest possible sound. After all, that is a good example of using only piano samples. what a digital instrument can do that acoustic instruments [great pipe organs the exception] do not. We also think of embodiment as a process. It is discovery, questioning and searching for a response in the context of an artistic discourse, an ongoing dialogue between the musician, the controller interface, the software mapping and the music.

7. FUTURE WORK Currently the instruments’ sensor implementations are limited by the small 7-bit MIDI controller resolutions. Although we have found ways around these limitations in regards to the control of audio and synthesis processing through data interpolation, averaging and smoothing, these Figure 6. Contra-monster techniques result in data hysteresis [sluggishness] and are really only a compromise. As our work developed our mapping strategies have grown much larger encompassing sets of over 50 parameters. These The next stage of development will involve upgrading the mappings are responding to the development of a reflexive instruments to OSC via Gluion interfaces. The Gluion is approach to playing and are designed to offer a large number capable of much faster scanning rates [up to 1ms] compared of possibilities to a performer. The central idea to these to other OSC interfaces on the market. This should offer a mappings is to create stock standard number of simultaneous sense of immediate control with a significant boost to sensors whose behavior can then be governed by a set of resolution. The signal latency of computer processing ancillary knobs or other more passive systems of control. remains a significant problem. We have also found software synthesizers to be limited in terms of polyphony and also in regards to tuning system implementation.

In conclusion, this project has created instruments, techniques and music exploring Grainger’s Free Music i.e. music using gliding tones and irregular rhythms. We have explored and extended Grainger’s ideas and legacy through the creation of new playable electroacoustic instruments. The research has created a folio of creative work including two finished CD albums, hours of recorded experimental work, International performances, exhibitions, television, radio performances and videos.

8. ACKNOWLEDGEMENTS We dearly acknowledge the contributions of three remarkable Figure 7. Visual Display Contra-monster Australians. Firstly to Garry Greenwood, who’s passing in April this year was a tragic loss. Garry’s incredible The Light-Harp for example has a standard playable set of leatherwork will inspire Australians for years to come. An sensors including light sensors [pitch or note/sample trigger], artist of rare talent we dedicate our work to Garry. We also breath control [attack/volume], two dimensional whammy- would like to acknowledge Jeff Pressing and his support and bar [push=feedback, side to side=delay time], a force belief in our work over the years previous to his passing. sensitive strip allows for another simultaneous control for Without Jeff’s support, we would never have believed in filtering, two large dials which can also be played controls developing our first prototype. Lastly we acknowledge David fine delay-time, modulation speed/depth and or specialized S. Brown, his friendship and remarkable talent for instrument filtering parameters. These main controls are supported by a innovation, skill and artistry, have given us great inspiration number of other controllers extending the mapping with up to from the beginning of our work in the early 1990’s. 64 additional transformations such as; transposition [+-6ve], re-scaling of temperament [quartertone, sixth-tone, 7-tet, 9- tet, 23-tet, 64-tet microtonal sets], fine tuning shifts, 7. REFERENCES modulation controls, envelope controls for filtering or [1] Burt, W. Percy Grainger’s Work with amplitude envelopes, signal processing parameters for delays, Melbourne Independent Media Centre, 2005 flangers, chorusing, flanging, distortion and granular effects. http://melbourne.indymedia.org/news/2005/02/87624.php Our experience has found that these larger mappings are intuitive; revealing more each time they are explored. [2] Cannon, C., Hughes, S., & Ómodhráin, S., “EpipE: Exploration of the Uilleean Pipes as a Potential Intrinsic to the success of mapping gesture is the notion of Controller for Computer-based Music” In the embodiment. This remains a subjective area of research and Proceedings of the 2003 Conference on New Interfaces we define embodiment as a convincing relationship between for Musical Expression (NIME03), Montreal, Canada physical gesture and resultant sound. Convincing in this sense

374 Proceedings of the 2006 International Conference on New Interfaces for Musical Expression (NIME06), Paris, France

[3] Choi, I., “A Component Model of Gestural Primitive [18] Ulyate, R., & Bianciardi. D., “The Interactive Dance Throughput” In the Proceedings of the 2003 Conference Club: Avoiding Chaos in a Multi Participant on New Interfaces for Musical Expression (NIME03), Environment.” In the Proceedings of the 2000 Montreal, Canada Conference on New Interfaces for Musical Expression, Seattle [4] Crab, S., The Free Music Machine http://www.obsolete.com/120_years/machines/free_mus [19] Young, D., Essl, G., “HyperPuja: A Tibetan Singing ic_machine/ Bowl Controller” In the Proceedings of the 2003 Conference on New Interfaces for Musical Expression [5] Dorum, E., Percy Grainger: the man behind the music (NIME03), Montreal, Canada published by IC & EE Dorum, Melbourne Australia, 1986

[6] Drefus, K., Interview with Burnett Cross, 5.3.1976 Audio collection Grainger Museum, 1976

[7] Giles, M., & Pear, D., “Grainger’s Top Ten” Review The Centre for Studies in Australian Music No. 15 June, 2002

[8] Kessous, L., & Arfib, D., “Bimanuality in Alternate Musical Instruments” In the Proceedings of the 2003 Conference on New Interfaces for Musical Expression (NIME03), Montreal, Canada

[9] Mulder, A., “Towards a choice of gestural constraints for instrumental performers” In Trends in Gestural Control of Music Ircam- Centre Pompidou, 2000

[10] Nelson, M., & Thom, B., “A Survey of Real-Time MIDI Performance” In the Proceedings of the 2004 Conference on New Interfaces for Musical Expression (NIME04), Hamamatsu, Japan

[11] Newton-Dunn, H., Nakano, H., & Gibson, J., “Block Jam: A Tangible Interface for Interactive Music” In the Proceedings of the 2003 Conference on New Interfaces for Musical Expression (NIME03), Montreal, Canada

[12] O’Modhráin, S., Essl, G., “PebbleBox and Crumble Bag: Tactile Interfaces for Granular Synthesis” In the Proceedings of the 2004 Conference on New Interfaces for Musical Expression (NIME04), Hamamatsu, Japan

[13] Paine, G., “Gesture and Music Interaction: Interactive Engagement Through Dynamic Morphology” In the Proceedings of the 2004 Conference on New Interfaces for Musical Expression (NIME04), Hamamatsu, Japan

[14] Pressing, J., “Cybernetic Issues in Interactive Performance Systems” Computer Music Journal, Vol. 14 No.1, 1990

[15] Rebelo, P., & Van Walstijn, “Designing Acoustic Thresholds” Les journées du design sonore, IRCAM 2004

[16] Shackel B. [1990] “Human Factors and Usability” In: Preece J. and L. Keller (Ed.) Human-Computer Interaction: Selected Readings. Prentice Hall 1990

[17] Siegel, J., & Siegel, W., “Categorical Perception of tonal intervals: Musicians can't tell sharp from flat.” Perception & Psychophysics, 399-407, 1977

375